Consciousness is one of the brain’s most enigmatic mysteries.
A new theory, inspired by thermodynamics, takes a high-level perspective of how neural networks in the brain transiently organize to give rise to memories, thought and consciousness.
The key to awareness is the ebb and flow of energy: when neurons functionally tag together to support information processing, their activity patterns synchronize like ocean waves.
This process is inherently guided by thermodynamic principles, which — like an invisible hand — promotes neural connections that favors conscious awareness.
Disruptions in this process breaks down communication between neural networks, giving rise to neurological disorders such as epilepsy, autism or schizophrenia.
By using thermodynamic principles, such as energy gradients, dissipation and approach to equilibrium, we have a way to start comprehending, or characterizing, how “mental things happen” and how they sometimes deviate towards neuropathological states, said study author Dr. Jose L. Perez Velazquez affiliated with the Ronin Institute in Montclair, NJ.
The results, funded by Novela Neurotech, were published in the open-access journal Frontiers in Neuroscience (On a Simple General Principle of Brain Organization).
Energy States as a Guiding Principle
Scientists have long hypothesized that consciousness arises from the coordinated activity among neurons, widely spread across the brain. One framework, the Global Workspace Theory, posits that some brain regions integrate information over space and time across a large number of connected brain areas, resulting in data that is globally available for diverse processes such as memory, attention and language.
Another hypothesis, the Integrated Information Theory, believes that consciousness is the result of heavily interconnected brain, the degree of which can be quantified.
Despite decades of work, these theories don’t directly tackle the harder question: what are the principles guiding these connections, so that consciousness arises in the brain?
As increasing effort focuses on communicating with locked-in patients and determining consciousness in intelligent machines, the pursuit of biological principles guiding brain organization becomes increasingly crucial.
The new work combines classical physics, especially some laws of thermodynamics, with modern recordings of neural activity, to paint a general framework of how changes in free energy — the amount of energy available inside a system — helps temporarily synchronize the activity in neural networks.
During conscious states, the brain has to actively integrate and segregate information from different senses and so consumes more energy than when unconscious.
Using available neural recordings from human participants during wakefulness, sleep, coma and seizures — each considered a brain “macrostate” — the team found that entropy during consciousness was higher than that during unconscious states.
As a concept, entropy can be interpreted and measured in many specific ways. Here, entropy is associated with the number of configurations of synchronized, or “connected”, brain networks.
“Energy is dissipated as more neurons become connected,” say Perez Velazquez and colleagues.
Models using thermodynamic equations show that healthy and conscious states have a tendency toward greater dissipation.
However, it is not just about how much free energy is in the brain.
Each macrostate is composed of multiple configurable microstates.
During conscious awareness, the brain has an optimal number of connected neural networks, and so many more microstates to support cognition.
In contrast, during unconscious states like seizures, there are too many connected neural networks resulting in fewer microstates — and so, lower entropy and higher free energy, causing the brain to malfunction.
“To maintain healthy brain states then is not about the total amount of energy in the brain […] but rather in how the energy is organized,” say the authors.
A General Principle of Brain Organization
Together, viewing brain organization through the lens energy gradients and dissipation combines into a theory — or tentatively, a “principle” — that can separate healthy, conscious brain states from unconscious ones.
The team thus believes that their approach can be used to further elucidate what happens when consciousness breaks, for example, in certain epileptic seizures.
Using the principle, the team offered an interpretation about how normal brain activity can transition into abnormal states.
When neurons hyperactivate, this results in higher-than-normal synchrony that either lasts too long or reaches too wide regions of the brain.
In other words, the brain settles on a state that is too stable. This idea agrees with a previous interpretation of consciousness, detailed in The Brain-Behavior Continuum, the subtle transition between sanity and insanity.
By using thermodynamic principles, such as energy gradients, dissipation and approach to equilibrium, we have a way to start comprehending, or characterizing, how “mental things happen” and how they sometimes deviate towards neuropathological states, said study author Dr. Jose L. Perez Velazquez affiliated with the Ronin Institute in Montclair, NJ.
As a result, the brain has lower entropy and so reduced ability to form variable brain activity patterns.
That is, it has fewer microstates, resulting in fewer configurations of interacting neural networks, which deprives the brain of its usual ability to quickly and flexibly adapt to the outside world. In some cases, consciousness also crumbles.
The team has now laid out experiments to test the theory. Dr. Diego M. Mateos at the Instituto de Matemática Aplicada del Litoral-CONICET-UNL in Santa Fe, Argentina, and Dr. Ramon Guevara Erra at the Laboratoire Psychologie de la Perception CNRS in Paris, France also contributed to the work.
The work is one project funded by Novela Neurotech, a brain-interface company based in Alameda, CA with roots in tackling epilepsy and interest in unveiling the neurobiological machinery behind consciousness.
“Novela is committed to collaborating with the neuroscience research community to illuminate the enduring mystery of consciousness. If we are to manage and cure neuropsychiatric disease, we must learn how the brain works, ” says Ray Iskander, CEO of Novela Neurotech.
A multitude of studies have focused on the investigation of patterns of correlated activity among brain cell ensembles based on magnitudes of a variety of synchrony indices or similar measures.
A prominent common aspect that is emerging from those studies is that of the importance of variability in the brain’s coordination dynamics.
In general, neurophysiological signals associated with normal cognition demonstrate fluctuating patterns of activity that represent interactions among cell networks distributed in the brain (Guevara Erra et al. 2016).
Similar result can be found in Werner (2009) and Pakhomov and Sudin (2013).
This variability allows for a wide range of configurations of connections among those networks exchanging information, supporting the flexibility needed to process sensory inputs.
Therefore, it has been argued that a certain degree of complexity in brain signals will be associated with healthy cognition, whereas low complexity may be a sign of pathologies (Garrett et al. 2013; Velazquez et al. 2003; Mateos et al. 2014; Dimitriadis et al. 2015). We sought to obtain evidence for the correlation between complexity in brain signals and conscious states, using brain electrophysiological recordings in conscious and unconscious states.
There exist a number of statistical measures to analyze electrophysiological recordings (Hlaváčková-Schindler et al. 2007).
In our work we use two well known measures, one statistical—Shannon entropy, a measure of unpredictability of information content in a message (Shannon 1948), and the other deterministic—Lempel–Ziv complexity, based on the minimum information required to recreate the original signal (Lempel and Ziv 1976).
For both measures, we use the quantifiers introduced by Bandt and Pompe (2002), called permutation vectors, which are based on the relationships of neighbor values belonging to a time series.
The Shannon entropy measure applied to the permutation vectors is known as permutation entropy (HPE) (Bandt and Pompe 2002). In a similar manner, the Lempel–Ziv complexity measure applied to the permutation vectors is called permutation Lempel–Ziv complexity (PLZC) (Zozor et al. 2014).
We used these two methods to obtain information about the signal’s dynamics from two different perspectives, probabilistic (HPE) and deterministic (PLZC).
The HPE and the LZC have been employed in previous studies analyzing electrophysiological recordings in epilepsy, coma or sleep stages (Olofsen et al. 2008; Ferlazzo et al. 2014; Nicolaou and Georgiou 2011; Casali et al. 2013; Zhang et al. 2001; Shalbaf et al. 2015). Moreover, there is an interesting relation, under certain restrictions, between Shannon entropy and Lempel–Ziv complexity that can naturally extend to HPE and PLZC (Cover and Thomas 2006; Zozor et al. 2014).
The results we obtain are shown in a complexity-entropy graphs. This kind of representation enables better visualization of the results giving a better understanding of the results especially for people who are not so familiar with these kind of analysis.
A recent study on chaotic maps and random sequences, it showed that the complexity-entropy graph allows for the distinction of different dynamics that was impossible to discern using each analysis separately (Mateos et al. 2017).
In our present work we analyze brain signals recorded using scalp electroencephalography (EEG), intracranial electroencephalography (iEEG) and magnetoencephalography (MEG), in fully alert states and in two conditions where consciousness is impaired: seizures and sleep.
The hypothesis derived from the previous considerations on variability of brain activity is that the brain tends towards larger complexity and entropy in wakefulness as compared to the altered states of consciousness.
Source:
Novela Neurotech
Media Contacts:
Press Team – Novela Neurotech
Image Source:
The image is in the public domain.
Original Research: Open access
“On a Simple General Principle of Brain Organization”. Jose L. Perez Velazquez, Diego M. Mateos and Ramon Guevara Erra.
Frontiers in Neuroscience doi:10.3389/fnins.2019.01106.
